Multi-unmanned aerial vehicle cooperative task allocation and path planning method based on genetic algorithm

Abstract

The invention discloses a multi-unmanned aerial vehicle cooperative task allocation and path planning method based on a genetic algorithm, and relates to an unmanned aerial vehicle path planning method. And operating a Dijkstra algorithm to accurately solve the shortest feasible path between all airports and task points, and finally generating a global path cost matrix for quick query. A complete and safe flight path that each unmanned aerial vehicle starts from an airport, sequentially accesses task points and returns is graphically displayed, a convergence curve of an algorithm, population diversity changes and performance comparison data under different task scales, unmanned aerial vehicle numbers and obstacle densities can be output, and a parameter sensitivity analysis function is supplemented. And the effectiveness, the stability and the practical value of the proposed method in a static urban environment are comprehensively verified.

Description

Technical Field

[0001] This invention relates to a method for planning unmanned aerial vehicle (UAV) missions, and more particularly to a method for multi-UAV collaborative mission allocation and path planning based on a genetic algorithm. Background Technology

[0002] In recent years, driven by technological advancements and policy support, the low-altitude economy has developed rapidly. The low-altitude economy generally refers to an economic system utilizing airspace below 1000 meters, relying on new types of aircraft such as drones and electric vertical takeoff and landing (eVTOL) aircraft, as well as business models such as aerial logistics and low-altitude tourism. National and local governments have successively introduced numerous supportive policies, conducted drone flight tests, and formulated airworthiness standards, aiming to promote the development of the low-altitude economy and enhance the intelligence level of the logistics industry.

[0003] In urban environments, drones have been increasingly used for delivery services, including express delivery, food delivery, and other small goods. With their advantages of efficiency, flexibility, and low cost, drones have brought revolutionary changes to traditional logistics and delivery. Beyond logistics, drones also show great potential in urban patrol and surveillance. Their high-altitude perspective and autonomous navigation capabilities allow them to cover a wider ground area, providing real-time video monitoring and information transmission.

[0004] However, deploying multiple drones simultaneously to perform various missions in the complex low-altitude urban environment presents numerous challenges. Limited low-altitude space in cities, coupled with tall buildings and no-fly zones, makes flight path planning and coordination extremely complex. Furthermore, different missions have varying requirements for flight time and routes; improper scheduling can easily lead to drone mission conflicts and overlapping flight paths, resulting in reduced resource utilization and low system efficiency.

[0005] Therefore, effectively solving the problem of task allocation for multi-UAV cooperatives has become a challenging research topic. Traditional methods are often inefficient when dealing with complex optimization problems involving multiple objectives and constraints, making it difficult to meet the needs of practical applications. Summary of the Invention

[0006] The purpose of this invention is to propose a multi-UAV cooperative task allocation and path planning method based on genetic algorithms. This method effectively solves the task allocation and path planning problem for multiple UAVs in complex urban environments through a reasonable encoding scheme and optimized genetic operator design.

[0007] This invention provides the following technical solution: A multi-UAV cooperative task allocation and path planning method based on genetic algorithms, the method comprising: S1 constructs a multi-UAV collaborative task allocation and path planning model based on urban low-altitude environmental information and UAV mission requirements. S2 constructs a multi-objective optimization function with the goals of maximizing mission completion value, minimizing total path cost, and balancing drone load. S3 sets the constraints for task allocation and path planning; S4 uses an integer sequence-based encoding mechanism to encode, decode, and evaluate the fitness of UAV mission sequences. S5 uses an improved genetic algorithm to solve the multi-UAV cooperative task allocation model and obtain the optimal task allocation scheme. S6 compares the performance of the improved genetic algorithm with the benchmark algorithm under different task sizes and environmental complexities.

[0008] The method for multi-UAV collaborative task allocation and path planning based on genetic algorithm is characterized in that the urban low-altitude environment information includes obstacle locations, no-fly zone ranges, and building distribution; the UAV task requirements include task point locations, task types, and task value; and the UAV information includes the number of UAVs, initial positions, and task execution capabilities.

[0009] The aforementioned multi-UAV cooperative task allocation and path planning method based on genetic algorithms, wherein the multi-objective optimization function is constructed with the objectives of maximizing task completion value, minimizing total path cost, and balancing UAV load, includes:

[0010] in, Represents the overall fitness function. Represents the total path cost. Indicates load balancing degree; This represents the importance coefficient of task j; Indicates drone Complete the task The probability of; Indicates drone Should the task be executed? , This represents the total flight distance of all drones; Indicates the maximum task execution time; Indicates drone The number of tasks; This indicates the average number of tasks.

[0011] The aforementioned multi-UAV cooperative task allocation and path planning method based on genetic algorithms includes setting constraints for task allocation and path planning, such as task allocation constraints, path feasibility constraints, and time coordination constraints. S31. Task allocation constraints:

[0012] Where N represents the total number of missions, M represents the total number of drones, and P represents the maximum number of missions per drone; S32. Route feasibility constraints:

[0013] in, Indicates drone To the mission distance, Indicates the minimum safe distance. Indicates the flight path, Indicates the area of ​​obstacles; S33. Task priority constraints:

[0014] in, Indicates drone Start executing the task Time, Indicates flight time, Indicates the task completion time. This represents the set of preceding tasks for task j.

[0015] The aforementioned multi-UAV cooperative task allocation and path planning method based on genetic algorithms includes an integer sequence-based encoding mechanism for encoding, decoding, and fitness evaluation of UAV task sequences, comprising: S41. A two-segment integer encoding scheme is adopted. The first segment represents the task execution order, and the second segment uses a separator to mark the UAV to which the task belongs. S42. Pre-calculate the shortest path between all key nodes using the visibility graph algorithm, and generate the path cost matrix; S43. During decoding, the chromosome sequence is mapped to the specific mission sequence and flight path of each UAV. S44. Fitness assessment comprehensively considers task value, path cost, time cost, and load balancing factors.

[0016] The aforementioned method for multi-UAV cooperative task allocation and path planning based on genetic algorithms includes the following steps: The improved genetic algorithm is used to solve the multi-UAV cooperative task allocation model. S51. Population initialization: An initial population is generated using a random uniform distribution strategy to ensure balanced task load. S52. Selection Operation: A tournament selection mechanism is adopted to balance individual advantages and population diversity. S53. Crossover operation: Use sequential single-point crossover to ensure the continuity of the task sequence; S54. Mutation operation: Adaptive uniform mutation is adopted, and the mutation probability is dynamically adjusted according to the iteration process. S54, Elite Retention: The top 10% of the best individuals in each generation are retained and directly enter the next generation; S54. Termination condition: The algorithm stops when the maximum number of iterations is reached or the fitness is stable.

[0017] The aforementioned multi-UAV cooperative task allocation and path planning method based on genetic algorithms includes the following adaptive uniform mutation operation: S541. Set the initial mutation probability to 0.05 and the maximum mutation probability to 0.3; S542. When the optimal fitness does not improve for 10 consecutive generations, the mutation probability will be increased by 50%. S543. The mutation operation adopts the task exchange method, randomly selecting two task positions to swap; S544. After mutation, a feasibility check is performed to ensure that the new individual meets all constraints.

[0018] The aforementioned multi-UAV cooperative task allocation and path planning method based on genetic algorithms includes performance comparisons under different task scales and environmental complexities, such as: S61. Set up three task scale scenarios: small scale (10 tasks), medium scale (20 tasks), and large scale (30 tasks). S62. Set two levels of environment complexity: simple environment (5 obstacles) and complex environment (15 obstacles); S63. Comparison metrics include: task completion rate, total path length, convergence algebra, and running time. S64. Comparative analysis with benchmark algorithms NSGA-II, particle swarm optimization, and simulated annealing.

[0019] The performance comparison analysis of the multi-UAV cooperative task allocation and path planning method based on genetic algorithm includes: S641. In all test scenarios, this method maintains a 100% task completion rate; S642. In terms of path optimization, this method reduces the path length by 15-25% compared to the baseline algorithm; S643. In terms of convergence speed, this method reduces the number of convergence algebras by 30-40% compared to the benchmark algorithm; S644. In terms of algorithm stability, this method demonstrates good robustness in problems of different scales.

[0020] The method for multi-UAV cooperative task allocation and path planning based on genetic algorithms further includes a visualization module for: S71. Real-time display of drone mission allocation results and flight paths; S72. Demonstrate the convergence process of the genetic algorithm and changes in population diversity; S73. Provide parameter sensitivity analysis and algorithm performance evaluation reports; S74 supports comparative analysis of simulation results under different scenarios.

[0021] The beneficial effects of the technical solution of this invention include at least the following: 1. This invention can effectively solve the problem of collaborative task allocation among multiple UAVs in complex urban environments, optimize flight paths while ensuring task completion rate, and reduce total system cost; 2. This invention improves the convergence speed and solution quality of the algorithm through a reasonable encoding scheme and genetic operator design; 3. This invention enhances the adaptability and robustness of the algorithm in different scenarios through an adaptive parameter adjustment mechanism; 4. This invention provides an efficient and reliable multi-UAV mission scheduling solution for application scenarios such as urban low-altitude logistics, intelligent delivery, and aerial security.

[0022] 5. This invention effectively solves the problem of task allocation and path planning for multiple UAVs in complex urban environments through a reasonable encoding scheme and optimized genetic operator design.

[0023] 6. This invention realizes a complete simulation process for multi-UAV collaborative task allocation and path planning. It begins with static environment modeling, which can read configuration files and construct an initial scene on a two-dimensional map, including polygonal obstacles, multiple task points, airports, and heterogeneous UAVs (with different task execution capabilities and success rates). Based on this, the system performs efficient path pre-calculation, uses the visibility graph method to connect all nodes without obstacles, and runs Dijkstra's algorithm to accurately solve for the shortest feasible path between all airports and task points. Finally, a global path cost matrix is ​​generated for quick querying. Subsequently, the software enters the core collaborative task allocation optimization stage, employing a genetic algorithm as the solution engine. By encoding the task allocation scheme into integer sequence chromosomes and undergoing an iterative evolutionary process including tournament selection, sequential crossover, adaptive mutation, and elite retention, a reasonable task execution sequence is dynamically planned for each UAV. The optimization objective is evaluated by a comprehensive fitness function, which simultaneously weighs multiple indicators such as task value, completion probability, total path cost, and maximum execution time, and imposes penalties for task conflicts or duplicate allocations. Thus, while ensuring a 100% task completion rate, the optimal or near-optimal scheduling scheme with the lowest total path cost and the most balanced resource allocation is found. Finally, the system provides comprehensive result visualization and performance analysis. It can not only graphically display the complete and safe flight trajectory of each UAV from the airport, sequentially visiting task points, and returning, but also output the algorithm's convergence curve, population diversity changes, and performance comparison data under different task scales, number of UAVs, and obstacle densities. With the addition of parameter sensitivity analysis, the effectiveness, stability, and practical value of the proposed method in a static urban environment are fully verified. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating the multi-UAV cooperative task allocation and path planning method based on genetic algorithms in an embodiment of the present invention. Figure 2 This is a schematic diagram of the urban environment map and obstacle modeling in an embodiment of the present invention; Figure 3 This is a schematic diagram of the visibility graph construction in an embodiment of the present invention; Figure 4 This is a schematic diagram of an integer sequence encoding scheme in an embodiment of the present invention; Figure 5 This is a schematic diagram of the iterative process of the genetic algorithm in an embodiment of the present invention; Figure 6 This is a schematic diagram of the path planning results in an embodiment of the present invention; Figure 7 This is a fitness change curve diagram in an embodiment of the present invention; Figure 8This is a graph showing the results of parameter sensitivity analysis in an embodiment of the present invention. Detailed Implementation

[0025] The present invention will now be described in further detail with reference to the embodiments shown in the accompanying drawings.

[0026] A multi-UAV cooperative task allocation and path planning method based on genetic algorithms, the method comprising: S1. Construct a multi-UAV collaborative task allocation model based on task scenario information, including UAV information, task point information, airport information, and obstacle information; S2. Construct a multi-objective optimization function with the goal of maximizing task completion value and minimizing total path cost; S3. Set constraints for task allocation and path planning; S4. Use an integer sequence encoding scheme to encode task allocation and execution order; S5. Design an improved genetic algorithm to solve the multi-UAV cooperative task allocation model; S6. Verify the effectiveness and performance of the algorithm through simulation experiments.

[0027] Furthermore, the task scenario information described in S1 includes: S11. The drone information includes the drone number, initial airport location, type of mission that can be performed, probability of mission completion, and flight speed. S12. Task point information includes task number, location coordinates, task type, importance, and completion time; S13. Airport information includes airport number, location coordinates, and capacity. S14. Obstacle information includes obstacle shape type, location coordinates, and size parameters. Further, the multi-objective optimization function described in S2 is: max F = V·R - C - T where: V is the total value of all completed tasks, calculated using the formula: V = ∑ᵢ∑ j 100·I j ·Pᵢ j R is the task completion rate reward / penalty factor. When all tasks are completed, R = 1.5; otherwise, R = c / m. C represents the total cost of the route, calculated as: C = 5.0·D_total; T represents the time cost, calculated using the formula: T = 2.0·T_max^0.7; where I j Pᵢ represents the importance of task j. jLet D_total represent the probability that drone i completes task j, D_total represent the total path length of all drones, and T_max represent the maximum time required for all drones to complete the task.

[0028] Furthermore, the constraints described in S3 include: S31. Task allocation constraint: Each task must be executed by one and only one drone. S32. Drone Capability Constraints: Each drone can only perform mission types within its capabilities. S33. Path feasibility constraints: All flight paths must avoid obstacles and no-fly zones; S34. Time Constraint: The total mission execution time for each drone cannot exceed its maximum endurance.

[0029] Furthermore, the integer sequence encoding scheme described in S4 is as follows: a one-dimensional integer list is used to represent the task execution order, the length of the list is equal to the total number of tasks, and each element corresponds to a task number; the task sequence is dynamically allocated to each UAV through the decoding process, and the least-task-first strategy is adopted during allocation.

[0030] Furthermore, the improved genetic algorithm described in S5 includes: S51. Employ a tournament selection mechanism to balance individual dominance and population diversity; S52. The sequential single-point crossover operator is used to ensure the continuity of the task sequence; S53. Adaptive uniform mutation operator is used to dynamically adjust mutation probability; S54. Introduce an elite retention mechanism to preserve the best individuals; S55. Employ a dynamic parameter adjustment mechanism to update crossover rate and mutation rate in real time based on population performance. Furthermore, the simulation experiments described in S6 include: S61. Conduct simulation experiments under different mission scales, number of drones, and obstacle density scenarios; S62. Evaluate performance metrics such as total path length, convergence algebra, runtime, and task completion rate. S63. Conduct sensitivity analysis on key parameters to determine the optimal parameter combination; S64. Compare and analyze the method of the present invention with the traditional task allocation method. Example

[0031] This invention provides a method for multi-UAV cooperative task allocation and path planning based on genetic algorithms, which can be implemented by electronic devices. Figure 1 This is a flowchart illustrating the process of this method. The processing flow includes the following steps: S1. Construct a multi-UAV collaborative task allocation model based on task scenario information; S2. Construct a multi-objective optimization function; S3. Set constraints for task allocation and path planning; S4. Use an integer sequence encoding scheme to encode task allocation and execution order; S5. Design an improved genetic algorithm to solve the model; S6. Verify the effectiveness and performance of the algorithm through simulation experiments.

[0032] The steps described above will be explained in detail below: First, step S1 will be described in detail: S11. Drone information includes drone number, initial airport location, executable mission type, mission completion probability, and flight speed. This information is loaded via configuration files to ensure that each drone has clearly defined attributes.

[0033] S12. Task point information includes task number, location coordinates, task type, importance, and completion time. Task points represent specific locations that the drone needs to access, such as delivery points or patrol points.

[0034] S13. Airport information includes airport number, location coordinates, and capacity. Airports are the starting and returning locations for drones, usually located at the edge of the map or in open areas.

[0035] S14. Obstacle information includes the obstacle's shape type, location coordinates, and size parameters. Obstacles are represented in polygon form and include non-flying areas such as buildings and no-fly zones.

[0036] Figure 2 The diagram shows a city environment map and obstacle model, clearly indicating the location and shape of the obstacles.

[0037] Next, step S2 will be described in detail: The multi-objective optimization function comprehensively considers task value, completion rate, path cost, and time cost: max F = V·R - C - T The calculation of each component is as described above. This function can complete more high-value tasks while minimizing path and time overhead, thereby improving the overall efficiency of the system.

[0038] A detailed description of step S3 is provided below: S31. Task assignment constraints ensure that each task is uniquely assigned, avoiding duplication or omission; S32. The drone capability constraint ensures that tasks are only assigned to drones with the corresponding capabilities. S33. Path feasibility constraints are implemented using a visibility graph algorithm to ensure that the path avoids all obstacles. S34. Time constraints take into account the actual endurance of the drone to ensure mission feasibility.

[0039] Figure 3 It demonstrates the process of constructing a visibility graph, forming a network of feasible paths by connecting nodes that are not blocked by obstacles.

[0040] A detailed description of step S4 is provided below: Integer sequence encoding uses a one-dimensional list of integers to represent the task execution order. For example, for 8 tasks, the chromosome might be represented as [3,1,5,2,4,6,8,7]. During decoding, a least-task-first strategy is used to dynamically allocate tasks to each UAV.

[0041] Figure 4 It shows a specific example of an encoding scheme, illustrating how to convert a task sequence into an actual task assignment.

[0042] A detailed description of step S5 is provided below: S51, the tournament selection mechanism randomly selects several individuals from the population, choosing the one with the highest fitness to enter the next generation, thus preserving both superior individuals and maintaining diversity; S52. Sequential single-point crossover is used to exchange parent gene segments at random positions to ensure the legitimacy of offspring. S53. Adaptive uniform mutation dynamically adjusts the mutation probability based on the convergence status, strengthening exploration in the early stage and enhancing development in the later stage. S54. The elite retention mechanism directly retains the best individuals in each generation to prevent the loss of excellent solutions; S55. Dynamic parameter adjustment updates algorithm parameters in real time based on population performance, improving search efficiency.

[0043] Figure 5 It demonstrates the iterative process of genetic algorithms, including operations such as selection, crossover, and mutation.

[0044] A detailed description of step S6 is provided below: S61. Conduct simulation experiments in different scenarios, including varying the task size (10-30 tasks), the number of drones (3-6), and the obstacle density; S62. Evaluate multiple performance indicators. Experimental results show that this method consistently maintains a 100% task completion rate, and the running time is acceptable. S63. Parameter sensitivity analysis determines the optimal parameter combination: population size 200, crossover rate 0.95, mutation rate 0.05, elite retention rate 0.1; S64. Compared with traditional methods, this method shows significant improvements in both path length and convergence speed.

[0045] Figure 6 The path planning results are displayed. Figure 7 The fitness change curve is shown. Figure 8 The results of the parameter sensitivity analysis are presented.

[0046] In an exemplary embodiment, a computer-readable storage medium is also provided, storing a computer program that, when executed by a processor, implements the above-described multi-UAV collaborative task allocation and path planning method.

[0047] The above description is merely a preferred embodiment of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention. To make the technical problems, technical solutions, and advantages of the present invention clearer, a detailed description will be provided below in conjunction with the accompanying drawings and specific embodiments.

[0048] This invention provides a method for multi-UAV cooperative task allocation and path planning based on genetic algorithms, which can be implemented by electronic devices. Figure 1 This is a flowchart illustrating the process of this method. The processing flow includes the following steps: S1. Construct a multi-UAV collaborative task allocation model based on task scenario information; S2. Construct a multi-objective optimization function; S3. Set constraints for task allocation and path planning; S4. Use an integer sequence encoding scheme to encode task allocation and execution order; S5. Design an improved genetic algorithm to solve the model; S6. Verify the effectiveness and performance of the algorithm through simulation experiments.

[0049] The steps described above will be explained in detail below: First, step S1 will be described in detail: S11. Drone information includes drone number, initial airport location, executable mission type, mission completion probability, and flight speed. This information is loaded via configuration files to ensure that each drone has clearly defined attributes.

[0050] S12. Task point information includes task number, location coordinates, task type, importance, and completion time. Task points represent specific locations that the drone needs to access, such as delivery points or patrol points.

[0051] S13. Airport information includes airport number, location coordinates, and capacity. Airports are the starting and returning locations for drones, usually located at the edge of the map or in open areas.

[0052] S14. Obstacle information includes the obstacle's shape type, location coordinates, and size parameters. Obstacles are represented in polygon form and include non-flying areas such as buildings and no-fly zones.

[0053] Figure 2 The diagram shows a city environment map and obstacle model, clearly indicating the location and shape of the obstacles.

[0054] Next, step S2 will be described in detail: The multi-objective optimization function comprehensively considers task value, completion rate, path cost, and time cost: max F = V·R - C - T The calculation of each component is as described above. This function can complete more high-value tasks while minimizing path and time overhead, thereby improving the overall efficiency of the system.

[0055] A detailed description of step S3 is provided below: S31. Task assignment constraints ensure that each task is uniquely assigned, avoiding duplication or omission; S32. The drone capability constraint ensures that tasks are only assigned to drones with the corresponding capabilities. S33. Path feasibility constraints are implemented using a visibility graph algorithm to ensure that the path avoids all obstacles. S34. Time constraints take into account the actual endurance of the drone to ensure mission feasibility.

[0056] Figure 3 It demonstrates the process of constructing a visibility graph, forming a network of feasible paths by connecting nodes that are not blocked by obstacles.

[0057] A detailed description of step S4 is provided below: Integer sequence encoding uses a one-dimensional list of integers to represent the task execution order. For example, for 8 tasks, the chromosome might be represented as [3,1,5,2,4,6,8,7]. During decoding, a least-task-first strategy is used to dynamically allocate tasks to each UAV.

[0058] Figure 4 It shows a specific example of an encoding scheme, illustrating how to convert a task sequence into an actual task assignment.

[0059] A detailed description of step S5 is provided below: S51, the tournament selection mechanism randomly selects several individuals from the population, choosing the one with the highest fitness to enter the next generation, thus preserving both superior individuals and maintaining diversity; S52. Sequential single-point crossover is used to exchange parent gene segments at random positions to ensure the legitimacy of offspring. S53. Adaptive uniform mutation dynamically adjusts the mutation probability based on the convergence status, strengthening exploration in the early stage and enhancing development in the later stage. S54. The elite retention mechanism directly retains the best individuals in each generation to prevent the loss of excellent solutions; S55. Dynamic parameter adjustment updates algorithm parameters in real time based on population performance, improving search efficiency.

[0060] Figure 5 It demonstrates the iterative process of genetic algorithms, including operations such as selection, crossover, and mutation.

[0061] A detailed description of step S6 is provided below: S61. Conduct simulation experiments in different scenarios, including varying the task size (10-30 tasks), the number of drones (3-6), and the obstacle density; S62. Evaluate multiple performance indicators. Experimental results show that this method consistently maintains a 100% task completion rate, and the running time is acceptable. S63. Parameter sensitivity analysis determines the optimal parameter combination: population size 200, crossover rate 0.95, mutation rate 0.05, elite retention rate 0.1; S64. Compared with traditional methods, this method shows significant improvements in both path length and convergence speed.

[0062] Figure 6 The path planning results are displayed. Figure 7 The fitness change curve is shown. Figure 8 The results of the parameter sensitivity analysis are presented.

[0063] In an exemplary embodiment, a computer-readable storage medium is also provided, storing a computer program that, when executed by a processor, implements the above-described multi-UAV collaborative task allocation and path planning method.

[0064] The above description is merely a preferred embodiment of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-unmanned aerial vehicle cooperative task allocation and path planning method based on a genetic algorithm, characterized in that, The method comprises: S1, constructing a multi-unmanned aerial vehicle (UAV) cooperative task allocation and path planning model according to urban low-altitude environment information and UAV task requirements; S2, constructing a multi-objective optimization function aiming at maximizing task completion value, minimizing total path cost and balancing UAV load; S3, setting constraint conditions for task allocation and path planning; S4, encoding, decoding and fitness evaluation of UAV task sequence based on integer sequence coding mechanism; S5, solving the multi-UAV cooperative task allocation model by using an improved genetic algorithm to obtain an optimal task allocation scheme; S6, comparing the performance of the improved genetic algorithm with that of a benchmark algorithm under different task scales and environment complexities.

2. The method of claim 1, wherein, The urban low-altitude environment information comprises obstacle position, no-fly zone range and building distribution; the UAV task requirements comprise task point position, task type and task value; and the UAV information comprises UAV number, initial position and task execution capability.

3. The method of claim 1, wherein, The multi-objective optimization function is constructed by maximizing the task completion value, minimizing the total path cost, and balancing the load of the unmanned aerial vehicle, and includes: ; wherein, denotes the comprehensive fitness function, denotes the total path cost, denotes the load balancing degree; denotes the importance coefficient of task j; denotes the probability that the UAV completes the task ; denotes whether the UAV executes the task , denotes the total flight distance of all UAVs; denotes the maximum task execution time; denotes the number of tasks of the UAV ; denotes the average number of tasks.

4. The method of claim 1, wherein, The constraint conditions for task allocation and path planning comprise task allocation constraint, path feasibility constraint and time coordination constraint: S31, task allocation constraint: ； wherein N represents total number of tasks, M represents total number of UAVs, and P represents maximum number of tasks for a single UAV; S32, path feasibility constraint: ； wherein, represents a drone to a task distance, represents a minimum safety distance, represents a flight path, represents an obstacle region; S33, priority constraint of tasks: ； wherein, represents a drone starts to perform a task the time, represents a flight time, represents a task completion time, represents a set of predecessor tasks of task j.

5. The method of claim 1, wherein, The integer sequence coding mechanism for encoding, decoding and fitness evaluation of UAV task sequence comprises: S41, adopting a two-section integer coding scheme, the first section representing task execution order and the second section representing UAVs to which tasks belong through a separator; S42, precomputing shortest paths between all key nodes by using a visibility graph algorithm to generate a path cost matrix; S43, mapping chromosome sequence to specific task sequence and flight path of each UAV during decoding; S44, comprehensively considering task value, path cost, time cost and load balancing factors for fitness evaluation.

6. The method of claim 1, wherein, The improved genetic algorithm for solving the multi-UAV cooperative task allocation model comprises: S51, population initialization: generating an initial population by using a random uniform distribution strategy to ensure task load balancing; S52, selection operation: adopting a tournament selection mechanism to balance individual advantage and population diversity; S53, crossover operation: adopting a sequential single-point crossover to ensure continuity of task sequence; S54, mutation operation: adopting an adaptive uniform mutation to dynamically adjust mutation probability according to iteration process; S54, elite reservation: reserving top 10% optimal individuals of each generation to directly enter next generation; S54, termination condition: stopping the algorithm when a maximum iteration number is reached or fitness is stable.

7. The method of claim 6, wherein, The adaptive uniform mutation operation comprises: S541, setting initial mutation probability to 0.05 and maximum mutation probability to 0.3; S542, increasing mutation probability by 50% when optimal fitness of 10 consecutive generations is not improved; S543, adopting a task exchange method for mutation operation, and randomly selecting two task positions for exchange; S544, performing feasibility check after mutation to ensure that the new individual meets all constraint conditions.

8. The method of claim 1, wherein, The performance comparison under different task scales and environment complexities comprises: S61, setting three task scale scenarios: small scale (10 tasks), medium scale (20 tasks), and large scale (30 tasks); S62, setting two environment complexities: simple environment (5 obstacles) and complex environment (15 obstacles); S63, comparison indexes comprise: task completion rate, total path length, convergence algebra, and running time; S64, comparison and analysis are performed with benchmark algorithms NSGA-II, particle swarm algorithm, and simulated annealing algorithm.

9. The method of claim 8, wherein, The performance comparison analysis comprises: S641, under all test scenarios, the method maintains 100% task completion rate; S642, in terms of path optimization, the method reduces path length by 15-25% compared with benchmark algorithms; S643, in terms of convergence speed, the method reduces convergence algebra by 30-40% compared with benchmark algorithms; S644, in terms of algorithm stability, the method shows good robustness in different scale problems.

10. The method of claim 1, wherein, The method further comprises a visual display module for: S71, real-time display of unmanned aerial vehicle task allocation results and flight path; S72, display of genetic algorithm convergence process and population diversity change; S73, provision of parameter sensitivity analysis and algorithm performance evaluation report; S74, support for simulation result comparison and analysis under different scenarios.

Images